|Publication number||US7098708 B2|
|Application number||US 11/186,451|
|Publication date||Aug 29, 2006|
|Filing date||Jul 20, 2005|
|Priority date||Oct 14, 2003|
|Also published as||US6958635, US20050077934, US20050253632, WO2005038636A1|
|Publication number||11186451, 186451, US 7098708 B2, US 7098708B2, US-B2-7098708, US7098708 B2, US7098708B2|
|Inventors||Amr M. Fahim|
|Original Assignee||Qualcomm, Incorporated|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (4), Referenced by (4), Classifications (25), Legal Events (2)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This application is a continuation of U.S. patent application Ser. No. 10/684,797, filed on Oct. 14, 2003, now U.S. Pat. No. 6,958,635.
The present invention relates generally to electronics circuits, and more specifically to a direct digital synthesizer (DDS).
In a modem communication device, multiple clock signals with frequencies that are unrelated may be needed for various functions. For example, a clock signal with a first frequency may be needed for a digital signal processing subsystem, another clock signal with a second frequency may be needed for a sampled analog subsystem, and so on.
Multiple clock signals with unrelated frequencies may be generated in various manners. In one conventional design, a clock generator with a single phase locked loop (PLL) is operated at a high frequency. The clock signal from this generator is divided in frequency by different integer values to obtain multiple output clock signals with different frequencies. This design places stringent requirements on the PLL in terms of performance and power consumption. In another conventional design, a separate PLL is provided for each subsystem requiring a clock signal with a different frequency. This design is undesirable because multiple PLLs for multiple clock signals normally consume a large amount of power and occupy a large area.
In yet another conventional design, an MN counter is used to divide an input clock signal (e.g., from a PLL) by a divider value to obtain an output clock signal with the desired frequency. The divider value is a ratio of two integer values M and N (i.e., N/M), where M<2·N for proper operation of the MN counter and N/M may be an integer or non-integer value. If the N/M divider value is not an integer, which is often the case, then the desired frequency is obtained by dividing the input clock signal in frequency by └N/M┘ for some of the time and by ┌N/M┐ for the remainder of the time, where └x┘ is a floor operator that provides the nearest lower integer value for x and ┌x┐ is a ceiling operator that provides the nearest higher integer value for x. This division with two integer values of └N/M┘ and ┌N/M┐ results in the output clock signal having inherent jitter that can be as large as one period of the input clock signal. For example, if the input clock frequency is 100 MHz, then the worst-case jitter for the output clock signal from the MN counter is 10 nsec.
Various methods for reducing jitter in the output clock signal from an MN counter have been proposed. For example, some methods reduce jitter by estimating the amount of phase shift needed in each output clock cycle to eliminate the jitter and then adjusting the output clock phase accordingly. In any case, most of these methods rely on absolute (voltage and/or current) reference levels to perform the phase shift estimation and/or adjustment and are thus prone to performance degradation due to circuit component mismatches and integrated circuit (IC) process variations.
There is therefore a need in the art for techniques to generate a clock signal having less jitter.
An MN counter with analog interpolation (referred to herein as an “MNA counter”) capable of generating an output clock signal having improved jitter performance is described herein. The jitter performance is minimally affected by IC process variations and system offsets using the design techniques described herein.
In a specific embodiment, the MNA counter includes an MN counter, a dither generator, an inverse unit, a multiplier, a delay generator, and a current generator. The dither generator provides a dither signal used to suppress spurious signals in the output clock signal caused by periodic jitter. The MN counter receives an input clock signal, the dither signal, and M and N values, accumulates M for each input clock cycle using a modulo-N accumulator, and provides an accumulator value and a counter signal. The counter signal has a frequency determined by the input clock frequency and the M and N values, and includes a pulse whenever the modulo-N accumulator wraps around. The inverse unit provides a Q value that is an inverse of M. The multiplier (which may be implemented with multiple pipelined stages to achieve higher operating speeds) multiplies the accumulator value with the Q value and provides an L-bit control signal. The current generator provides a reference current for the delay generator. The delay generator receives the counter signal and the L-bit control signal, compares a differential signal generated based on the counter and control signals, and provides the output clock signal. The leading edges of the output clock signal have variable delay determined by the L-bit control signal and the reference current.
The delay generator may be implemented with a differential design that utilizes two banks of capacitors. The capacitors in each bank may be implemented with binary decoding or thermal decoding and are selectable by the L-bit control signal. The selected capacitors in one bank are charged by one current source, and the selected capacitors in the other bank are discharged by another current source. The differential signal is defined by the two voltages on the selected capacitors in the two banks. The variable delay is determined by the amount of time taken for the two voltages to cross each other. Details of the delay generator are described below.
The current generator may be implemented with a replica delay generator and a current locked loop. The replica delay generator has the same design as the delay generator and is configured to provide a predetermined amount of delay (e.g., one half input clock period of delay) when the proper reference current is received. The current locked loop adjusts the reference current so that the predetermined amount of delay is obtained. The capacitors for the replica delay generator are matched to the capacitors for the delay generator, and the capacitors for both delay generators are arranged in a two-dimensional array using a common centroid layout to achieve good matching. Low-power is achieved by enabling the replica delay generator a sufficient number of (e.g., two) input clock cycles prior to each output clock edge transition, then disabling the replica delay generator after the transition is complete.
Various aspects, embodiments, and features of the invention are described in further detail below.
The features and nature of the present invention will become more apparent from the detailed description set forth below when taken in conjunction with the drawings in which like reference characters identify correspondingly throughout and wherein:
The word “exemplary” is used herein to mean “serving as an example, instance, or illustration.” Any embodiment or design described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments or designs.
MN counter 220 operates as follows. Register 222, summers 224 and 226, and multiplexer 228 collectively implement a modulo-N accumulator that stores a value ranging from 0 to N−1. For each input clock cycle, the accumulator accumulates M with the current accumulator value and provides the first combined value, which is stored back in register 222 if an overflow has not occurred. An overflow occurs whenever the first combined value exceeds N and is indicated by the OVFb signal being at logic low. When an overflow occurs, N is subtracted from the first combined value and the result is stored in register 222. A pulse is provided on the output clock signal when an overflow occurs.
An MN counter with analog interpolation (an “MNA counter”) can be used to reduce jitter and spurs. The MNA counter attempts to reduce jitter by shifting the position of the leading edges (e.g., rising edges) of the output clock signal such that all output clock periods are the same. This is achieved by determining the amount of phase shift needed for each output clock cycle to obtain the desired output clock period and then advancing the leading edge accordingly.
where ACCi is the accumulator value at the time of overflow. The leading edge of the output clock is advanced by TCLKin/3 when the accumulator value is one, advanced by 2TCLKin/3 when the accumulator value is two, and not advanced when the accumulator value is zero.
If the phase shift can be generated exactly and if the leading edges can be advanced by this phase shift without errors, then all of the output clock cycles will have equal period and the ideal MNA counter will have zero jitter. A delay generator can be used to generate the desired phase shift for the MNA counter. The delay generator can be designed to generate phase shifts in discrete steps. Higher accuracy can be attained for the delay generator with greater circuit complexity, more die area, and higher power consumption. Thus, there is a trade-off between the accuracy of the delay generator and other system considerations.
For a delay generator with L-bit accuracy, where L>1, the required phase shift can be estimated as:
A phase shift to advance the output clock edge (i.e., a negative phase shift) can be obtained by operating the delay generator one input clock cycle early and generating a delay that is complementary to the negative phase shift. The delay may be expressed as:
In general, for an L-bit delay generator, the worst-case cycle-to-cycle jitter is TCLKin/2L, which is a reduction by a factor of 2L over the jitter generated by the MN counter. Jitter is thus exponentially reduced for larger values of L (assuming no degradation due to circuit implementation). However, circuit complexity, area, and power consumption also increase exponentially with L. A suitable choice for L can be determined based on jitter requirements and other factors.
Dither generator 440 generates the dither signal and is described below. An inverse unit 450 receives M, derives an inverse of M, and provides an inverse value (Q), where Q≅1/M. A multiplier 460 receives the ACC value, the Q value, and the Ce2 signal. Multiplier 460 multiplies the ACC value with the Q value, when enabled by the Ce2 signal, and provides a P value for an L-bit control signal. The P value corresponds to the term (1−Fi/2L) in equation (4). The P value is obtained by quantizing the product of ACCi and Q using L bits (with rounding for the least significant bit) and inverting all of the L bits. The P value is indicative of the amount of delay (if any) required for the current output clock cycle. A delay generator 470 receives the P value and the Ce1 and Ce2 signals and generates the output clock signal (CLKout). The output clock signal has each leading edge shifted by the delay indicated by the P value. A current generator 480 generates the reference currents, Irefp and Irefn, for delay generator 470. Each of the units in MNA counter 400 is described in further detail below.
Within MN counter 420, a carry-save adder (CSA) 524 receives and combines M, the dither signal, and the accumulator value (ACC) (i.e., three input terms) and provides two output terms. A CSA 526 receives and sums the two output terms from CSA 524 and subtracts N and provides two output terms. A carry lookahead adder (CLA) 528 a combines the two output terms from CSA 524 and provides the first combined value (V1) to a multiplexer 530. A CLA 528 b combines the two output terms from CSA 526 and provides the second combined value (V2) to multiplexer 530. CLA 528 b also provides the Ce2 signal. Multiplexer 530 provides either the V1 or V2 value to a register 522 depending on the Ce2 signal. MN counter 420 can be operated at a high operating speed because of the use of carry-save adders and carry lookahead adders. However, other designs may also be used for MN counter 420.
Dither generator 440 provides the one-bit dither signal that is used to randomize the jitter and reduce the amplitude of the spurs caused by periodic jitter from MN counter 420. The dither signal can be generated based on a pseudo-random number (PN) sequence. Dither generator 440 may thus be implemented with a linear feedback shift register (LFSR) that is configured to implement a polynomial generator for a PN sequence. For example, a 26-bit LFSR that implements a polynomial generator x25+x24+x20+1 may be used for dither generator 440. The dither signal comprises a repeating pseudo-random sequence of +1 and −1 and does not introduce an average frequency offset to the output clock frequency. Other designs for generating the one-bit dither signal may be used without affecting the scope of the embodiments herein.
Inverse unit 450 generates a value of Q≅1/M. Inverse unit 450 may be implemented with a serial division algorithm, a look-up table, or some other manner. Since the Q value is typically computed once and does not change for a given operating mode, this value may be provided by a unit external to MNA counter 400. For example, a controller can compute and provide the Q value via a register.
For the first pipeline stage, latches 612 a and 612 b latch the ACC and Q values, respectively, with the Ce2 signal. Latches 612 a and 612 b inhibit the ACC value from rippling through multiplier 460 if no overflow occurs in MN counter 420. A partial product tree generator 614 performs multiplication of the ACC and Q values by generating partial product terms and accumulating these terms in multiple accumulation stages. Partial product tree generator 614 provides two partial product terms.
For the second pipeline stage, D flip-flops 622 a and 622 b store the two partial product terms from partial product tree generator 614. A carry-select adder 624 combines the two partial product terms from D flip-flops 622 a and 622 b and provides the final result. The partial product accumulation is divided into two pipeline stages to support higher operating speed for multiplier 460. The multiplication can be partitioned into more than two stages for even higher operating speed.
For the third pipeline stage, a D flip-flop 632 a stores the most significant bits (MSB) of the final result from carry-select adder 624 and a D flip-flop 632 b stores the least significant bit (LSB) of the final result. An AND gate 634 performs a logical AND of the LSB from D flip-flop 632 b with a Round Enable signal. A CSA 636 combines the MSB of the final result from D flip-flop 632 a with the output of AND gate 634 and provides the combined result to a D flip-flop 638. The rounding of the final result can reduce truncation errors by ½ LSB. D flip-flop 638 provides the P value, which is approximately equal to ACC/M and quantized to L bits. For illustrative ease, the bit inversion to obtain the term (1−Fi/2L) in equation (4) is not explicitly shown in
A delay generator can generate a number of discrete delays by charging and discharging a bank of capacitors. Different delays can be obtained by turning on (i.e., selecting) different combination of capacitors in the bank. The delay generated by the delay generator can be expressed as:
where Cload is the load capacitance (which is dominated by the selected capacitors in the bank), Ich is the current used to charge and discharge the load capacitor, and Vswing is the voltage swing of the delay generator. Equation (5) shows that the delay is dependent on three parameters that are in turn dependent on the IC manufacturing process. Since accurate generation of the delay is needed to achieve good jitter performance, techniques are described herein to mitigate the effects of Cload, Vswing, and Ich on jitter performance.
Delay generator 470 includes two banks of capacitors, 710 and 720. Bank 710 includes S capacitors 712 a through 712 s, each having one end coupled to circuit ground and the other end coupled to a node Vp via switches 714 a through 714 s, respectively. S is dependent on the number of bits (L) and the design for delay generator 470. Bank 720 includes S capacitors 722 a through 722 s, each having one end coupled to circuit ground and the other end coupled to a node Vn via switches 724 a through 724 s, respectively. A switch 732 has one end coupled to node Vp and the other end coupled to circuit ground. A switch 734 has one end coupled to node Vp and the other end coupled to one end of a current source 736. The other end of current source 736 couples to a supply voltage (VDD). A switch 742 has one end coupled to node Vn and the other end coupled to VDD. A switch 744 has one end coupled to node Vn and the other end coupled to one end of a current source 746. The other end of current source 746 couples to circuit ground.
A comparator 750 has a non-inverting input coupled to node Vp, an inverting input coupled to node Vn, and an output that provides the output clock signal. A decoder 770 receives the P value from multiplier 460 and generates control signals for switches 714 and 724 to select the desired capacitors and deselect the remaining capacitors.
Current source 736 receives the reference current Irefp from current generator 480 provides a charging current of Idg. Current source 746 receives the reference current Irefn from current generator 480 and provides a discharging current of Idg. Current sources 736 and 746 may be implemented with constant-gm current sources or some other types of current sources. Capacitors 712 and 722, switches 714 and 724, and decoder 770 may be implemented as described below.
Delay generator 470 operates as follows. When the Ce2 signal is activated due to an overflow in MN counter 420, switches 714, 724, 732, and 742 are all turned on for one input clock cycle, capacitors 712 in bank 710 are discharged to circuit ground by switch 732, capacitors 722 in bank 720 are precharged to VDD by switch 742, node Vp is at circuit ground, node Vn is at VDD, and the output clock signal is at logic low. One input clock cycle later, switches 732 and 742 are turned off and only selected ones of switches 714 and 724 in banks 710 and 720 are turned on by the P value from multiplier 460. The selected capacitors in banks 710 and 720 are those with their associated switches turned on. In this same input clock cycle, switches 734 and 744 are turned on by the Ce1 signal, the selected capacitors in bank 710 are charged toward VDD by current source 736, the selected capacitors in bank 720 are discharged toward circuit ground by current source 746, the voltage on node Vp rises, and the voltage on node Vn drops. A differential signal is defined by the voltages on nodes Vp and Vn. When the voltage on node Vp exceeds the voltage on node Vn (i.e., when the two voltages cross), the output clock signal transitions to logic high. The leading edge of the output clock signal is thus determined by the amount of delay provided by delay generator 470. The enable signals for delay generator 470 are delayed appropriately to line up with the arrival of the P signal from multiplier 460.
The output clock signal provided by delay generator 470 within MNA counter 400 does not have 50% duty cycle. To obtain an output clock signal with 50% duty cycle, the M value may be doubled, and the output signal from comparator 750 may be divided by two (2) to obtain the output clock signal. The maximum N/M ratio is then limited to 0.5 for proper operation of the MNA counter.
The delay generated by delay generator 470 may be expressed as:
where Cunit is a unit capacitance. In equation (6), the factor VDD/2 is the voltage swing for the capacitors in each bank with the differential design and corresponds to Vswing in equation (5). The factor 2L·Cunit is the total capacitance for all capacitors in one bank. The factor [1−ACCi/M] corresponds to the P value from multiplier 460. The factor Cdg=2L·Cunit·[1−ACCi/M] is the capacitance for all selected capacitors in one bank and corresponds to Cload in equation (5). The unit capacitance Cunit and the charging current Idg are selected such that delay generator 470 provides one input clock period (TCLKin) of delay when all capacitors in the bank are selected (i.e., when Cdg=2L·Cunit). The capacitor size directly affects jitter, area, and power consumption. A suitable capacitor size can be selected based on a tradeoff of all of these considerations.
The accuracy of delay generator 470 is dependent on the accuracy of the unit capacitance Cunit, the voltage swing VDD, and the charging current Idg, which can vary due to IC process variations. A current locked loop can be used to generate a requirement current that tracks process variation of the delay generator circuitry and mitigates the effects of these three parameters on the accuracy of delay generator 470.
A phase-frequency detector (PFD) 830 compares the phase of the reference clock signal and the phase of the delayed clock signal and provides a phase error. PFD 830 may be implemented with an early-late detector that is known in the art. A charge pump 832 converts the phase error into a current. A loop filter, implemented with a single capacitor 834, filters the current from charge pump 832 and also converts the current into a voltage. Capacitor 834 can be a small capacitor if the input clock rate is high. A voltage-to-current (V-to-I) converter 840 converts the voltage on capacitor 834 back into a current using current mirrors and provides the reference currents Irefp and Irefn to replica delay generator 870. Replica delay generator 870 adjusts its delay based on the reference currents Irefp and Irefn such that the delayed clock signal is time-aligned with the reference clock signal. V-to-I converter 840 also provides the reference currents Irefp and Irefn to current sources 736 and 746 within delay generator 470.
The closed loop transfer function H(s) for current locked loop 480 a may be expressed as:
where Icp is the current of charge pump 832, Clf is the capacitance of capacitor 834 for the loop filter, Kv2i is the gain of V-to-I converter 840 (in units of μA/V), and Kdg is the gain of replica delay generator 870 (in units of nsec/μA). Equation (7) indicates that current locked loop 480 a is a single pole system that is unconditionally stable. However, similar to a delay locked loop, the current locked loop can false lock to a zero time period or a double time period. False lock can be prevented by ensuring that the forward gain is not too large. One method of achieving this is to control the charge pump current Icp (e.g., from 5 μA to 40 μA, in 5 μA steps). A large charge pump current Icp may be used initially to achieve fast locking. A small current may be used thereafter to prevent the current locked loop from overshooting. The gain Kv2i is adjusted such that delay generator 870 is capable of producing one half input clock period of delay for the entire range of input clock frequencies and over all process corners. The gain Kdg is controlled by the size of the capacitors in delay generator 870.
Replica delay generator 870 has the same design as delay generator 470. The delay generated by replica delay generator 870 may be expressed as:
where Crdg is the capacitance and Irdg is the charging current for replica delay generator 870. Replica delay generator 870 is designed to provide a delay of TCLKin/2 with capacitance Crdg and current Irdg. Delay generator 470 is designed to provide a delay of TCLKin with capacitance 2L·Cunit and current Idg. The capacitance to current (C/I) ratio for replica delay generator 870 is thus one half of the C/I ratio for delay generator 470, since the same voltage swing VDD is used for both generators. For example, the same charging current may be used for both delay generators (i.e., Idg=Irdg), and the capacitance of replica delay generator 870 may be set to half of the total capacitance of delay generator 470 (i.e., Crdg=2L·Cunit/2). In this case, the accuracy of delay generator 470 is dependent on the matching between the capacitance of delay generator 470 and the capacitance of replica delay generator 870. Current locked loop 480 a thus removes the effects of the supply voltage VDD and the charging current Idg, and the accuracy of delay generator 470 is not impacted by these two parameters. Current locked loop 480 a also effectively removes the effect of the load capacitor Cdg, and the accuracy of delay generator 470 is dependent on the matching of the capacitors for generators 470 and 870 instead of the capacitance of these capacitors.
The capacitors for delay generator 470 and replica delay generator 870 may be implemented in various manners. In one embodiment, binary decoding is used for the capacitors in each of banks 710 and 720 in
With the current locked loop, the accuracy of delay generator 470 is dependent on the matching between the capacitors for delay generator 470 and replica delay generator 870. To achieve good matching, the capacitors for both delay generators 470 and 870 can be implemented as one array arranged in a two-dimensional (2-D) common centroid layout.
For the exemplary implementation shown in
The MNA counter design described herein provides good performance and other advantages. The differential design allows for a reduction in the size of the capacitors in delay generators 470 and 870. The smaller capacitor size results in less power being consumed to generate the desired delay. Moreover, a smaller area is required to implement the smaller-size capacitors. The current locked loop removes the dependency on voltage swing and charging current, which improves the accuracy of the delay generator. To reduce power consumption, the current locked loop and multiplier may be turned on one or two input clock cycles before they are needed (e.g., using the Ce1 and Ce2 signals) and turned off thereafter.
For the receive path, signals transmitted by base stations are received by an antenna 1212, routed through a duplexer (D) 1214, and provided to a receiver unit (RCVR) 1216. Receiver unit 1216 conditions (e.g., filters, amplifies, and frequency downconverts) the received signal and digitizes the conditioned signal to provide samples, which are provided to a digital signal processor (DSP) 1220 for further processing. For the transmit path, data to be transmitted from wireless device 1200 is provided by DSP 1220 to a transmitter unit (TMTR) 1218. Transmitter unit 1218 conditions (e.g., filters, amplifies, and frequency upconverts) the data and generates a modulated signal, which is routed through duplexer 1214 and transmitted via antenna 1212 to the base stations.
DSP 1220 includes various processing units such as, for example, an internal controller 1222, a processor 1224, a memory unit 1226, a bus control unit 1228, and an audio processor 1236, all of which are coupled via a bus 1238. DSP 1220 further includes a PLL 1230 that receives a reference signal (e.g., from a temperature compensated crystal oscillator (TCXO)) and generates a master clock for DSP 1220. PLL 1230 may generate various clock signals (e.g., by dividing the master clock with different integer values) for the processing units within DSP 1220 and possibly for processing units external to DSP 1220 (e.g., a main controller 1240 and a main memory unit 1242). An MNA counter 1234 receives the master clock and generates a first low-jitter clock signal for audio processor 1236. An MNA counter 1234 receives the master clock and generates a second low-jitter clock signal for another processing unit (e.g., analog-to-digital converters within receiver unit 1216). The first and second low-jitter clock signals have different frequencies. In general, one MNA counter may be used to generate each different clock frequency that is not an integer multiple of the master clock. MNA counters 1232 and 1234 may each be implemented with MNA counter 400 in
The direct digital synthesizer with analog interpolation (i.e., the MNA counter) described herein may be implemented in an application specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a processor, a controller, a micro-controller, a microprocessor, and other electronic units. The MNA counter may be implemented within the one or multiple integrated circuit (IC) dies and in the one or multiple ICs. For example, all units of the MNA counter may be implemented on one IC die. As another example, the digital portion of the MNA counter (e.g., MN counter 420, dither generator 440, inverse unit 450, multiplier 460, and D flip-flops 422 and 424 in
The MNA counter may also be fabricated with various IC process technologies such as CMOS, NMOS, BJT, and so on. The MNA counter may also be fabricated using different device size technologies (e.g., 0.13 mm, 30 nm, and so on).
Portions of the MNA counter (e.g., inverse unit 450 and multiplier 460) may be implemented in software. For a software implementation, the modules (e.g., procedures, functions, and so on) may be used to perform some of the functions described herein. The software codes may be stored in a memory unit (e.g., memory unit 1226 or 1242 in
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US6359950 *||Mar 5, 2001||Mar 19, 2002||Infineon Technologies.||Digital PLL (phase-locked loop) frequency synthesizer|
|US6583674 *||Mar 9, 2001||Jun 24, 2003||Nokia Mobile Phones Ltd||Fractional multimodulus prescaler|
|US6731176 *||Mar 13, 2002||May 4, 2004||Atheros Communications, Inc.||Synthesizer with lock detector, lock algorithm, extended range VCO, and a simplified dual modulus divider|
|US6836526 *||Feb 25, 2003||Dec 28, 2004||Agency For Science, Technology And Research||Fractional-N synthesizer with two control words|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US7667504 *||Mar 11, 2008||Feb 23, 2010||International Business Machines Corporation||Signal delay element, method and integrated circuit device for frequency adjustment of electronic signals|
|US7764208 *||Nov 21, 2008||Jul 27, 2010||Stmicroelectronics S.R.L.||Clock dithering process for reducing electromagnetic interference in D/A converters and apparatus for carrying out such process|
|US20090021288 *||Mar 11, 2008||Jan 22, 2009||International Business Machines Corporation||Signal Delay Element, Method and Integrated Circuit Device for Frequency Adjustment of Electronic Signals|
|US20090140896 *||Nov 21, 2008||Jun 4, 2009||Stmicroelectronics S.R.L.||Clock dithering process for reducing electromagnetic interference in d/a converters and apparatus for carrying out such process|
|U.S. Classification||327/155, 377/48, 377/47, 327/160|
|International Classification||H03K5/00, H03K5/135, G06F1/08, G06F1/02, H03L7/081, H03L7/06, H03K23/68, H03K3/017, H03K5/156|
|Cooperative Classification||H03K5/1565, G06F1/022, H03K2005/00071, H03L7/0814, G06F2211/902, H03K2005/00032, H03K5/135, G06F1/08|
|European Classification||H03K5/156D, H03K5/135, H03L7/081A1, G06F1/02W|
|Jan 22, 2010||FPAY||Fee payment|
Year of fee payment: 4
|Jan 28, 2014||FPAY||Fee payment|
Year of fee payment: 8